Anyone know anything about 3d engraving files?

I was going to post this in the free design category, where the tutorials get posted, but there’s supposed to be a file included with every new thread. So I guess here? If this belongs somewhere else please feel free to move it to a new home! So, just to be careful, I’m not looking for free designs and I guess you aren’t allowed to post files as examples. We’ll just have to use our words.

Internet at large is telling me greyscale TIFF files, 300dpi or greater. Is that right? Or is any raster file acceptable? Do they need to be greyscale as in a setting for the file, or as in just use black and white within the normal rgb or cmyk whatever setting (sorry if that doesn’t make a ton of sense, I’m not very familiar with this)? Let’s say you engrave the same design 1/8 inch deep and 1/4 inch deep. Will the images be the same just shallower, or will the 1/8 inch engraving just not have the deepest/darkest values? By which I mean, is ‘black’ always ‘deepest’, or does it correlate to some sort of fixed depth. Does that make sense anywhere except my head? Also, does it actually need to be smooth gradients, or can you just have different shades of grey with hard edges, each denoting a different depth (presumably making a choppier, less smooth engrave)?

It seems like an awesome feature… that I have zero idea how to take advantage of or make designs for :slight_smile: If there’s a site or tutorial that answers all my dumb questions, feel free to just tell me to shut up and point the way! I’ve tried googling but I am clearly not using the right words :blush:


Never done it, but yeah, eventually I want to know too.

Did you catch the thread that @tom did here?

There are a couple of other guys on there, @mdcoley5 and @lairdknox who might know a bit more about it too, from lithopane work on 3D printers. (Starts with a gradient depth map.)

Those three might be our experts.


I did see that thread originally, but missed the mentions about depth maps and lithopane. That sounds hard. Lol.

Thanks for the info – I’ll check it out more.

1 Like

My understanding is you specify the depth for black and it scales in between. So the 1/8" version will have all the levels in the file, but each half as deep as the 1/4" version.


I don’t know how useful this will be, but I used the search phrase “creating a greyscale depthmap in photoshop” on youtube and got a bunch of hits, mainly geared towards CNC routing, many of which touch on file setup. If you have time you could repeat that search and look further. From what I understand at this point, we get to skip most of the painful stuff they show(ie tool paths, g-code setup, etc.)!

here are a couple that could be good for getting the gears turning, using photoshop’s 3D features to render/preview.

(edit- note that we won’t have to convert our images to “3d” per se, but we could preview what the results might be when we engrave)


Bookmarked for after the holidays! :smile:


Here’s a manual way to get it done from existing pictures. Takes a little work, but looks pretty simple.


Both bookmarked :relaxed:

I started out with the intent on creating lithophanes on a desktop cnc. In order to get a better understanding of how a cnc works I began engraving photographs on small planks of wood. WIthin the first few months of having my cnc, I purchased several 3D printers. I felt a 3D printer would be a more efficient way of doing lithophanes. I soon realized the 3D printer was really good (IMHO) at making lithophanes that I never attempted this with my cnc.

Through my research, I purchased several programs that have assisted me in creating gcode out of photographs. I never really thought about the differences in using a grayscale photo versus a color photo. So, tonight, I ran several test files using two of my programs. The results were very similar, or the same, despite being a color or grayscale photo. Now, one might assume that each program converts the image to grayscale prior to analyzing it and creating the g code. This may be true, however, this doesn’t require any editing or photo manipulation on my part. The program just works :slight_smile:

Each program is different in how they approach depth settings. One program works based off of the z height dimension I enter. And another program allows me to enter any number I want for both the white and black settings. I believe this is more of a scale than a unit of measurement. For example, the difference between 1.25 and 3 is grossly different in the way the depth looks in the preview screen. Both of these programs are designed for cnc.

I have a third program specifically made for engraving photographs with a diode laser. Unfortunately there is no preview window to see the processed file looks. The programs accepts color or grayscale photographs. Looking at the user guide it states, “will generate a gcode file from five popular image file types,
either color or gray scale.”

So there you have it. My experience comes from programs that do the conversion for me. However, photographs are certainly different than vector files.

@scatterbrains, thank you for posting that great tutorial.


Would you mind sharing the names of the software you use?

1 Like


Meshcam (my first choice for most cnc projects)
PicEngrave Pro5
Microcarve V4
Lithophane 3D (

I also use a program called imagetogcode. Unfortunately this is no longer available but, it does has some nice features.

I have to admit, I’m becoming very fond of Easel from Inventables.


Not trying to squirt lemonade in anyones’ Cheerios, but I’m fairly confident the three workflow examples provided so far aren’t worth pursuing if the intent is to achieve what we’ve all seen in 3D laser examples in the other threads.

A greyscale depth-map may be known by different names. In 3D sculpting, it may be referred to as an Alpha… which itself denotes something else to Photoshop jockeys. In the way the Glowforge and sculptors use a Depth-map/Alpha, black defines the bottommost end of the pool and white defines the untouched water surface. How far the Glowforge burns into the material when it encounters black might be a function of how much power the laser puts out and/or how long it lingers on that spot. Correlate this with your material HEIGHT and TYPE, and it should be the way you’d control the max limits of how far the laser drills down. If I had a pre-pro unit, it’d probably spend the next 6 months 3D burning various types of materials, cutting the cross section to observe the depth characteristics, and THEN testing various gradient curve levels to find the optimal curve for each specific kind of wood species, acrylic, etc.

I’m not at all convinced the MasterFX tutorial provided a true depth map and highly doubt that workflow translates in any way to something more fully 3D; face, sphere, vase, etc.

The tutorial generated something 3D-ish, but I’m still failing to see its applicability. Bottom line? A traditional 2D grayscale photo (from a camera) does not have the kind of information that will ever translate into a proper 3D object.

The Triaxes site took something that was 2D photo-based and attempted to fudge its way into becoming a 3D depth map. I too was curious to see exactly what the result would be like. (Well, I kinda already envisioned) I realized I could use ZBrush to bring this around full-circle. Borrowing their depth maps as an Alpha, I simply setup ZBrush to “carve” into a virtual block in vagely the same way the laser cutter would do it – that is, penetrate deepest with the black, for white pixels leave the surface alone, and with the grays in-between treat accordingly. If these results are acceptable to you, carry on.

Next up, testing the notion that any ol’ 2D photo can be fed into this process to get a full 3D object out of it…

Facing head-on, the engrave has the tonal values that mimic the image used.

But just a slight tilt, and the 3D-ness that everyone expects is shown to be an epic fail. The eyebrows and nose undercut are sunk way in, The rest of the face isn’t extruding out in the proper bas-relief manner.

A depth map derived from a 3D file is the only way to carry the information through. 3D model it, 3D sculpt it, 3D scan it. Any of these three need to be in the equation.

Going back to a previous thread, I took one of my depth maps derived from a 3D CAD model and mimicked a laser engrave…

The low-resolution depth map originally derived from a 3D mesh scan

Not terribly shabby for a garbage-resolution input depth-map Alpha measuring 475x384. A 1200 pixel higher-rez version along with a nudge on the Alpha blur slider would’ve done the trick.

As of now, the easiest way of generating a Depth Map from an existing 3D object is through ZBrush as described here. I could tap the GrabDoc button all day long to get exactly the type of images the Glowforge needs.

I know some may be asking, so I checked: Now that the cheaper ($150) ZBrushCore is out, can we make DepthMaps/Alphas in that? Short answer. Nope. After scouring a few threads and 3 ZBrushCore tutorial clips, they clearly stripped out the GrabDoc button found in the Alpha selection palette of ZBrush 4R7.

Those who have the time to push many more buttons to get a depth map can try Blender’s method:

Apologies to those who wished to finish out 2016 without having to gaze at Blender. :smiling_imp:


This is also very helpful - I’ll be looking at 3D mapping options too for true gradient _shaping_*. Thanks!

* (perhaps not using Blender though…chuckle!):smile:

It’s fantastic that there appear to be a couple of different options for people with different design comfort levels.

Thanks for letting us know you’re familiar with this part…I’ll be pestering you later. :wink: :innocent:


This is what I was concerned about.

And I’m way less excited about 3d engraving now. Lol.


I’m not. I’m less concerned with the way things look in models and simulations than they do in real life materials I’ll be using.

It is true that we’re taking a 2D image and making it into something that is not quite what we could get from a 3D image but that’s okay. 3D movies aren’t really 3 dimensional, they just appear that way. Similarly the GF translation into a “varying depth engrave” of a 2D gray-scale image appears to be 3D - at least enough for me. What I’ve seen is pretty awesome so far. Not like I wander the world with 3D scanning/capture hardware.


Oh, I just meant I’m less excited about it personally. The results that they’ve gotten on the glowforge are really, really beautiful, but it would appear the files to make that happen are way more complicated than I’m interested in exploring :slight_smile:

1 Like

Nope. The GF examples were from sending it a gray-scale photo (one mouse click operation in any photo or design program) and sending it to the GF.

But the limitation is that the black is the deepest part of the engrave, the white unengaged and everything in between at depths relative to their darkness.

With Rhino or Blender or other tools and a 3d scan capture of something there are other tools (CNC) that can make a “more” 3D engrave.

If you like what GF has shown than you don’t need to worry about all that other stuff - 3D engraves are a mouse click away :smiling_face:

The GF magic is doing all the stuff the other posters were talking about (or at least enough for the results to be what we’ve seen) without us having to do the work. That’s seriously cool.


Wait. So it doesn’t need a depth map thing?

If I send this to the glowforge (which is just a circle and a square with a gradient fill, saved as a png)

it would have enough info to 3d engrave that?

Or no? I’m so confused.


Yes it will do a 3D engrave. The point is just sending a normal photograph puts the dark areas at the back. While that might be optically correct the 3D contour in no way relates to depth in the scene.

Unlike most people here I wont’ be using my Glowforge to make art as I haven’t got an an artistic bone in my body (I might use other people’s arty designs for gifts, etc). My main use will be making machines and mechanisms. For that I will design everything in OpenSCAD which can output 2D SVG or DXF or 3D STL files.

It seems Glowforge won’t accept proper 3D files, despite being marketed as a 3D printer, but I can easily write a program to convert STL to a PNG depth map and will share it here if somebody doesn’t beat me to it. It remains to be seen how accurate the depth will be in acrylic and Delrin.


That will work for a 3D engrave.

A long time ago when those cross your eyes and a 3d something will pop out pictures were popular I learned how to make depth maps in Photoshop using the gradient tool. They weren’t anything amazing, but over time I got good at using the lighting/darkening brushes to paint 3D depth maps and make them into stereoscopic images.

It’s pretty fun. Isn’t hard to do. There are techniques you can use to create some cool stuff. Like paste a pattern into Photoshop with a 25% gray as the background. Like a lattice or something. Then you can paint stuff on top of it in a lighter shade of gray to make it show up in front. Etc etc